Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
@ConfusedAnimator7pls read my comment lol I said I don’t do art as much as I use…
ytr_UgxS7lqu2…
G
Me - Ask Robot one if he thinks the guy in the stupid hat thinks he looks cool? …
ytc_Ugz7zVgL9…
G
I don’t mean to change subject, but I feel when personal anecdotes become involv…
rdc_dzp9m1g
G
Thank you for sharing your perspective! On AITube, we explore the capabilities o…
ytr_UgylVjvv4…
G
skill issue mate, coding jobs aint going anywhere look at the corporations which…
ytc_UgzUjrMem…
G
Hi Padmapriya, you got the right answer. Kudos.
The contest is over and winners …
ytr_Ugw_cqqrQ…
G
I agree ai should be stopped now. We don't need it. It makes us lazy. It's to…
ytc_Ugz4xGXFs…
G
So these AI companies risk destroying humanity and Trump's AI czar wont do anyth…
ytc_Ugwzuknd8…
Comment
1\2 BULL CRAP. ** PLEASE! IS THE CHAT EXACT VERSION MENTIONED OR NOT
ChatGPT said:
The exact chat transcript described in the video is not real and is not the version published in the medical case.
Here is what is known from reliable sources only (the Annals case report + news outlets summarizing it):
What the real case report says
The clinicians did not have access to the patient’s original chat logs.
The patient could not recall his exact prompts because he had been using the AI for months.
The clinicians themselves tested ChatGPT by asking similar questions and found that it could suggest bromide as a chloride-reducing alternative.
The case report does not print any chat transcript.
No version of ChatGPT (“4,” “5,” etc.) and no specific dialogue appears in the clinical publication.
What news reports say
News coverage (Guardian, LiveScience, etc.) confirms the same point:
no chat log was available; only reconstruction from patient history and clinician testing.
Reporters clearly state that the exact prompts and responses remain unknown.
What the video adds
The video includes:
A supposedly detailed interaction with “ChatGPT 5” in November 2025
Direct quotes like “You are correct!” or “There isn’t actually any guy.”
AI defensiveness or semantic arguments
Repeated “bromide warning” messages
None of this appears in the case report.
None of it is documented by any reliable external source.
These dialogues are invented narrative for storytelling.
The chat in the video is not the exact version from the real case.
The actual case has no preserved transcript, and the video’s quoted conversations are fictionalized and not from any published medical or technical record.
youtube
AI Harm Incident
2025-11-25T23:0…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | distributed |
| Reasoning | mixed |
| Policy | unclear |
| Emotion | outrage |
| Coded at | 2026-04-27T06:24:53.388235 |
Raw LLM Response
[
{"id":"ytc_UgwA6Z9vzFT38LVnwrJ4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"approval"},
{"id":"ytc_UgwYASihMkcAsTuK3nZ4AaABAg","responsibility":"ai_itself","reasoning":"unclear","policy":"unclear","emotion":"indifference"},
{"id":"ytc_UgzF52w-c0ldD1IU7Bl4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"approval"},
{"id":"ytc_UgxogMPl1bVgwqf9AfN4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"unclear","emotion":"fear"},
{"id":"ytc_UgwEMFaFSOBeQil7EZR4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"unclear","emotion":"indifference"},
{"id":"ytc_Ugxwhgz6YFsCPf5Sljp4AaABAg","responsibility":"distributed","reasoning":"mixed","policy":"unclear","emotion":"outrage"},
{"id":"ytc_UgxZME2ZkUYDFg_DvjB4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"mixed"},
{"id":"ytc_UgwNOMiwsGRlbJO8A2V4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"approval"},
{"id":"ytc_UgwKx1AgrZYPdax1ORl4AaABAg","responsibility":"company","reasoning":"deontological","policy":"unclear","emotion":"fear"},
{"id":"ytc_UgwDWz3-YwOqEy67TQF4AaABAg","responsibility":"distributed","reasoning":"consequentialist","policy":"none","emotion":"approval"}
]